The sum in the second case is, of course, not so obvious as it is in the of 2 plus 2. But again, this makes no difference. To be analytic it is not required that it be obvious What obvious to one person is not obvious to another, and what is obvious to a person at one time may not be at another time. What may not be obvious to you and me may well be obvious to a mathematical genius. Obviousness is a psychological characteristic that is i no way involved in the conception of being analytic. Propositions of arithmetic are analytic because their denial is self-contradictory, whether the self-contradictoriness is immediately obvious or not. To a being with very great mathematical powers the sum of very large numbers would be as obvious as "2 +2= 4" is to us.
3. "But the meaning of the two is not the same: 40694 and 27593 are not a part of the meaning of "68287.' When you ask me what I mean by this number, I don't give the other two-or any of the other sets of numbers that when added together would yield it. So how can the statement be analytic if the one is not all or even a part of the meaning of the other? But it doesn't need to be a part of the meaning, in the sense of what we mean when we say it. A may be B although "B" may not be what we mean when we say "A." 68287 may not be what we mean by 40694 and 27493 but it is the sum of those two numbers just the same. It is still a necessary truth, and the denial of it would still be self-contradictory.
4. "But isn't even such a simple statement plus 2 equals 4' a generalization from experience? Don't we learn its truth from experience? And isn't it based on instances? I first learn about 2 and 2 houses, then about 2 and 2 apples, and so on. How is its being learned from experience compatible with its being analytic?'
Of course I learn that 2 and 2 makes 4, and probably we all learned it as children, using examples such as houses and apples. But what is it that we learned? Is it about apples and houses? No, it is simply that 2 and 2 when added together makes 4; all the business about houses and apples was just window-dressing. What we learned was that the symbol "4" is equivalent in meaning to the symbol "2 and 2" that these two expressions can be used interchangeably.
We do indeed learn the meanings of words through experience-how else? But this does not have anything to do with whether the propositions in which they occur are analytic. What makes them analytic is whether or not their negation self-contradictory. To say that 2 and 2 does not make 4 would be is to say that and 1 and 1 does not make 1 and 1 and 1 and 1 which is self-contradictory.
When we put two pennies into our new piggy bank, and later two more pennies, we learned to say that we had put in four pen simply because putting 4 pennies in" means the same as "putting 2 pennies and 2 pennies in." We learned to say it as a result of our experience our experience of learning language but what we said was a necessary truth, and analytic. B we also learned to predict that if we should open the bank later we would find four pennies in it. In this case what we learned was not an arithmetical truth but a truth about the world that we might call the conservation of pennies. and, unlike "2 plus 2 equals 4," this proposition might have turned out to be false without contradiction. If it had turned out to be false, we could still agree that "2 and 2 makes 4" is an analytic truth following from the definition of what we mean by "2 "4," plus and "equals.